Medical errors are a paramount concern in health care, and often they are a byproduct of ineffective processes and systems within hospitals, operating rooms, and clinics. The recent Ebola case in Dallas is a stark reminder of that fact. On October 8 at Texas Health Presbyterian Hospital, Thomas Eric Duncan became the first person in the U.S. to die from the Ebola virus. Despite informing an Emergency Room (ER) nurse that he had recently traveled from Liberia, Mr. Duncan was sent home with a presumed viral illness after his initial presentation to Texas Presbyterian. The attending ER physician never became aware of this recent travel, for his workflow in the electronic health record (EHR) was separate from that of the nurse.
Texas Presbyterian released a statement on October 2, acknowledging there was a miscommunication between the physician and nurse due in part to the design of the EHR. Since then, two nurses involved in Mr. Duncan's care contracted (and have recovered from) the Ebola virus themselves, and outlets like Vox have questioned if the CDC's Ebola treatment protocols were too complicated for proper adherence.
From my experience having worked in over fifteen different hospitals/clinics since beginning medical school, errors of communication and complication are far from rare. We only heard so much about the Dallas Ebola case because of its high profile and the gravity of the virus. This past summer, the office of Senator Barbara Boxer (D-CA) published a report on medical errors in America. The report cites a 2013 Journal of Patient Safety article that estimates between 210,000 and 440,000 Americans die every year due to medical errors and other preventable harm at hospitals -- or the equivalent of a jumbo jet crashing every day with no survivors. By that count, medical errors can be considered the nation's third-leading cause of death, trailing only heart disease and cancer.
As the Dallas Ebola case highlights, one major issue underlying medical errors is that many practices and workflows in health care are overly complicated, clumsy, and broadly inefficient. Why would a nursing workflow with vital patient information not be designed to show up in the physician's workflow (and vice versa)? And why would that vital information not have been flagged so it was essentially impossible to ignore?
At present, providers and other health care staff often have to track down information in several disparate locations -- sometimes even within the same chart. That process is quite time-consuming, often leading to missed information and redundancy of testing/efforts. For example, the EHR at my current hospital requires me to look at more than 10 different screens per patient to get all the daily information I need to write a single note. When one multiplies that number by 15 to 20 patient encounters per day, there is ample opportunity to overlook key data by accident.
Similarly, clinical processes and patient handoffs are generally disjointed and unnecessarily complicated. Just this past week, I admitted a patient from the ER. Both the ER nurse and physician notes contained only the names of his medications without dosages. By the time I went to examine the patient, his medication list was nowhere to be found, he didn't know his medications, and his physician's office and pharmacy were already closed for the day.
Those examples are just two of many systems-based challenges that medical providers face consistently. The more salient point is that existing health care systems/processes contain an excess of touch-points where it is too easy to make a mistake -- often unintentionally. In my opinion, the goal should be to design systems that reduce the number of these touch-points. For those that remain, we should aim to simplify roles and clarify responsibilities in order to minimize the likelihood of potential error. Almost every day in the hospital, I find myself asking, "If I could redesign this system to make it more simple and transparent, what would it look like?" That question is a valuable one physicians, nurses, administrators, and operations staff should ask routinely.
The prevalent culture in health care is to create workarounds when complications arise rather than fixing the root problem. Issues are addressed on an ad hoc, individual basis, over time leading to patchwork processes without consideration for the overall system. Instead, I would argue health care needs more standardization, automation, and evidence-based protocols. Common processes in health care should function very much like an assembly line. Each step should be clearly defined with appropriate resources available, and every participant should have well understood responsibilities.
Years ago, Dr. Atul Gawande popularized the simple concept of using checklists in health care, which are broadly applicable beyond the intensive care unit (ICU) or operating room yet still not universal. And medicine could learn much from operations in other industries, such as Toyota and its famed Toyota Production System (TPS). Recent automobile recalls aside, TPS is regarded as one of the seminal collections of operating philosophy and practice. Toyota used that system to revolutionize auto manufacturing, which is why TPS is taught in business schools all over the world. The hospital is not a classic factory and patients are more complicated than cars, but concepts like continuous improvement, iterative problem solving, process efficiency, and proactive employee engagement are quite relevant to medical practice.
Furthermore, it is no secret that the use of technology in health care processes (as opposed to diagnostic and therapeutic usages) is lacking. Even when technology is present, it is often used ineffectively. In a recent study conducted at the University of California-San Francisco (UCSF), five ICUs collectively had more than 2.5 million patient alarms over a 31-day span. That is more than 82,500 alarms per day! Within that cohort, there were more than 1.1 million alarms for irregular heart rhythms, of which almost 89 percent were false positives. If roughly nine out of 10 alarms are inappropriate, how is a nurse or doctor supposed to know which patient alarms deserve his/her attention? Clearly, technology alone does not fix health care. But there is no doubt that the thoughtful application of technology has a major role in health care delivery moving forward. That paradigm is one reason there has been such aggressive growth in digital health investment over the past few years.
Every field has areas of inefficiency, but health care may be worse than most. And unlike most other fields, the ramifications involve personal safety and quality of health. Patients and the public usually only hear about mistakes when they lead to harm, which luckily is a much smaller subset of total mistakes. However, Sen. Boxer's office points out that mistakes are still far too common, and the recent Dallas Ebola case has brought health care operations to the forefront of media coverage.
Now is the time for medical communities everywhere to examine existing processes critically, pursue thoughtful advances in how we deliver care, and promote a culture that engages staff in the improvement process. Taking care of patients is not only about the therapies we provide but also having the most effective care delivery systems possible. By that metric, American health care still has significant room to grow.